日志 7.7.8.2008

工作报告    董迎松   引言：如下介绍的是蛋白质2级结构预测软件下载、安装、运行及其结果。   1、Jnet  version：alpha 0.1 1）A Neural Network Protein Secondary Structure Prediction Method. Application of enhanced multiple sequence alignment profiles to improve protein secondary structure prediction .By Cuff J. A and Barton G.J. 2) 下载地址：http://www.compbio.dundee.ac.uk/jpred_v2/jnet/ 3）安装：jnet.linux.tar.gz         cp jnet.linux.tar.gz  ./jnet        #上传         tar zxvf jnet.linux.tar.gz          #解压               cd ./jnetsrc/src                 make               cp jnet ../bin               export JNET=installdirectory/jnet   #设置环境变量   4) 运行：         ./bin/jnet -p ./sequence.fas   5) 结果： Length = 76  Homologues =  1  RES         : MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQQRLIFAGKQLEDGRTLSDYNIQKES  ALIGN    : ---EEEE----EEEEE------HHHHHHHHH--------HHHHHHHH------------HHHHHH  CONF      : 04576650684357740687605466766520588887346777630122377654541111454  FINAL     : ---EEEE----EEEEE------HHHHHHHHH--------HHHHHHHH------------HHHHHH  SOL25     :                                                                   SOL5       :                                                                    SOL0       :                                                                     RES         : TLHLVLRLRGG  ALIGN    : HHHHHHH----  CONF      : 33322200389  FINAL     : HHHHHHH----  SOL25     :             SOL5       :             SOL0       :                             6）批量预测：   /*        jnet.cpp        By Dong Yingson        Function: using jnet to make predication of protein's secondary structure. Jnet will be called in this program.        Input: protein.txt        Output:jnet_result.txt, ss.txt       call: jnet */ #include <fstream>   using namespace std; int main() {        char in_string[50000];        int count=1;                      ifstream protein("protein.txt");        ofstream jnet_input;        //ifstream jnet_output;        ofstream jnet_result("jnet_result.txt");               protein>>'>'>>count>>endl>>in_string;          while(!protein.eof())        {               count++;               jnet_input.open("jnet_input.fasta");               jnet_input<<'>'<<count<<endl<<in_string<<endl;               jnet_input.close();                 system("./bin/jnet -p ./jnet_input.fasta>>jnet_result.txt");              protein>>'>'>>count>>endl>>in_string;        }          protein.close();        jnet_result.close();        }//int main()     编译jnet.cpp:     gcc –o jnet_prediction jnet.cpp     运行cpp文件：        ./jnet_prediction     or   nohup ./jnet_prediction &  7) 结果处理：提取预测结果中的结构信息     2、sspord   version 4.01 1）SSprod 4 is the first hybrid approach of combining neural network (ab initio) and homology analysis to improve the prediction of secondary structure and solvent accessibility. 2）下载：http://download.igb.uci.edu/ 3) 安装： tar xzf sspro4.tar.gz    #解压 cd sspro4 open configure.pl and set the $install_dir to the sspro4 installation dir.  #修改安装程序 ./configure.pl         #运行安装程序 4）运行： ./bin/predict_acc.sh  sspro4_input.fasta  sspro4_output.txt 批量预测： /*        sol_sspro4.cpp        By Dong YS        Function: using sspro4 to make predication of protein's solvent accessibility.        sspro4 will be called in this program.        Input: protein.txt        Output:sspro4_result_sol.txt, solvent.txt       call: sspro4 */ #include <stdio.h> #include <fstream> #include <iostream> using namespace std;   int main() {        string in_string;        int count=1;           ifstream protein("protein.txt");        ofstream sspro4_input;        ifstream sspro4_output;        ofstream sspro4_result("sspro4_result_sol.txt");               protein>>in_string;                            cout<<count;               sspro4_input.open("sspro4_input.fasta");               sspro4_input<<'>'<<count<<endl<<in_string<<endl;               sspro4_input.close();                 system("../bin/predict_acc.sh sspro4_input.fasta sspro4_output.txt");                                     sspro4_output.open("sspro4_output.txt");               sspro4_output>>in_string;                 sspro4_output>>in_string;               sspro4_result<<"> "<<count<<endl;               sspro4_result<<in_string<<endl;               sspro4_output>>in_string;               sspro4_result<<in_string<<endl;                             sspro4_output.clear();               sspro4_output.close();                 protein>>in_string;               count++;        }          protein.close();        sspro4_result.close();        }//int main()  while(!protein.eof())        {               sspro4_input.open("sspro4_input.fasta");               sspro4_input<<'>'<<count<<endl<<in_string<<endl;               sspro4_input.close();                 system("../bin/predict_acc.sh sspro4_input.fasta sspro4_output.txt");                                     sspro4_output.open("sspro4_output.txt");               sspro4_output>>in_string;                 sspro4_output>>in_string;               sspro4_result<<"> "<<count<<endl;               sspro4_result<<in_string<<endl;               sspro4_output>>in_string;               sspro4_result<<in_string<<endl;                             sspro4_output.clear();               sspro4_output.close();                 protein>>in_string;               count++;        }          protein.close();        sspro4_result.close();        }//int main()       编译：sol_sspro4.cpp     gcc –o sspro sol_sspro4.cpp 运行cpp文件：        ./sspro     or   nohup ./sspro &    5) 结果处理:   3、CLC protein workbench  version 3.6.2   1) 基于windows   2）安装：双击，根据提示安装   3）运行：      （1）导入数据   图1  clcbio示意图    左上角Import导入数据，以序列1UBQ为例，如图所示。（2）结构预测单击左下角Toolbox中Protein Analyses的 Predict Secondary structure。既开始预测。（3）结果导出    图1左上角Export，并选择csv存储 Start;"End";"Region name";"Note 1" 3;"8";"beta strand";"Predicted by CLC Protein Workbench version 3 12;"16";"beta strand";"Predicted by CLC Protein Workbench version 3 23;"32";"alpha helix";"Predicted by CLC Protein Workbench version 3 40;"46";"alpha helix";"Predicted by CLC Protein Workbench version 3 48;"50";"alpha helix";"Predicted by CLC Protein Workbench version 3 67;"73";"alpha helix";"Predicted by CLC Protein Workbench version 3    再将结果转换成序列形式：   CEEEEEECCCE EEEEEEECCC CHHHHHHHHH HHCCCCCCCE EEEECCCEEE